Based on the operational regional ensemble prediction system (REPS) in China Meteorological Administration (CMA), this paper carried out comparison of two initial condition perturbation methods: an ensemble transform Kalman filter (ETKF) and a dynamical downscaling of global ensemble perturbations. One month consecutive tests are implemented to evaluate the performance of both methods in the operational REPS environment. The perturbation characteristics are analyzed and ensemble forecast verifications are conducted; furthermore, a TC case is investigated. The main conclusions are as follows: the ETKF perturbations contain more power at small scales while the ones derived from downscaling contain more power at large scales, and the relative difference of the two types of perturbations on scales become smaller with forecast lead time. The growth of downscaling perturbations is more remarkable, and the downscaling perturbations have larger magnitude than ETKF perturbations at all forecast lead times. However, the ETKF perturbation variance can represent the forecast error variance better than downscaling. Ensemble forecast verification shows slightly higher skill of downscaling ensemble over ETKF ensemble. A TC case study indicates that the overall performance of the two systems are quite similar despite the slightly smaller error of DOWN ensemble than ETKF ensemble at long range forecast lead times.

The regional ensemble prediction system (REPS) of North China is currently under development at the Institute of Urban Meteorology, China Meteorological Administration, with initial condition perturbations provided by global ensemble dynamical downscaling. To improve the performance of the REPS, a comparison of two initial condition perturbation methods is conducted in this paper: (i) Breeding, which was specifically designed for the REPS, and (ii) Dynamical downscaling. Consecutive tests were implemented to evaluate the performances of both methods in the operational REPS environment. The perturbation characteristics were analyzed, and ensemble forecast verifications were conducted. Furthermore, a heavy precipitation case was investigated. The main conclusions are as follows: the Breeding perturbations were more powerful at small scales, while the downscaling perturbations were more powerful at large scales; the difference between the two perturbation types gradually decreased with the forecast lead time. The downscaling perturbation growth was more remarkable than that of the Breeding perturbations at short forecast lead times, while the perturbation magnitudes of both schemes were similar for long-range forecasts. However, the Breeding perturbations contained more abundant small-scale components than downscaling for the short-range forecasts. The ensemble forecast verification indicated a slightly better downscaling ensemble performance than that of the Breeding ensemble. A precipitation case study indicated that the Breeding ensemble performance was better than that of downscaling, particularly in terms of location and strength of the precipitation forecast.

In this study, based on the retrieval data from the Fengyun geostationary meteorological satellite and the Tropical Rainfall Measuring Mission satellite, a large-scale precipitation case in eastern China is selected to address the systematic deviations of deterministic forecasts for clouds and precipitation. A multi-physical ensemble transform Kalman filter (ETKF) is constructed in this research based on the Weather Research and Forecast model version 3.6, and its forecasting ability in terms of cloud-top height and temperature, hydrometeors, and precipitation is evaluated by quantitatively comparing three microphysical parameterization schemes (Lin, Morrison, and CAM5.1 schemes) and their corresponding multi-physical ensemble mean. The results show that the Lin, Morrison, and CAM5.1 schemes all underestimate the range of cloud systems and have different advantages and disadvantages in forecasting different elements, while the forecasting improvement of the multi-physical ensemble mean is limited. However, the multi-physical ETKF can effectively improve the forecast accuracy of the cloud system range. In addition, the multi-physical ETKF has the advantages of different physical parameterization schemes, which can dramatically improve the forecast accuracy of cloud hydrometeors, reduce precipitation forecast errors, and improve threat scores.

To improve the skills of the regional ensemble forecast system (REFS), a modified ensemble transform Kalman filter (ETKF) initial perturbation strategy was developed. First, sensitivity tests were conducted to investigate the influence of the perturbation scale on the ensemble spread growth and forecast skill. In addition, the scale characteristic of the forecast error was analyzed based on the results of these tests, and a new initial condition perturbation method was developed through scale-selection of the ETKF perturbations, namely, ETKF-SS (scale-selective ETKF). The performances of the ETKF-SS scheme and the original ETKF (hereinafter referred to as ETKF) scheme were tested and compared. The results showed that the large-scale perturbations were much easier to grow than the original ETKF perturbations. In addition, scale analysis of the forecast error showed that the large-scale errors showed significant growth at the upper levels, while the small and meso-scale errors grew fast at the lower levels. The comparison results of the ETKF-SS and the ETKF showed that the ETKF-SS perturbations had more obvious growth than the ETKF perturbations, and the ETKF-SS perturbations in the short-term forecast lead times were more precise than the ETKF perturbations. The ensemble forecast verification results showed that the ETKF-SS ensemble had a larger spread and smaller root mean square error than the ETKF at short forecast lead times, while the probabilistic scores of the ETKF-SS also outperformed those of the ETKF method. In addition, the ETKF-SS ensemble can provide a better precipitation forecast than the ETKF.

With the development of technology, major operational centers and scientific research institutions have carried out research and development of convection permitting ensemble prediction technology and systems. In order to meet the needs of accurate forecasting in the capital, the Convection permitting ensemble prediction system of North China CMA-BJ-EN v1.0 developed by the Institute of Urban Meteorology was officially put into operation in January 2023. This paper introduces the research and development background and key technologies of the system in detail. The results show that the CMA-BJ-EN v1.0 system has a resolution of 3km convective resolution scale, covering 21 members, and can provide hourly probability prediction results of 48h forecast effectiveness in North China; The system is coupled with advanced technologies such as initial conditon perturbation of ensemble data assimilation, stochastic physical process tendency(SPPT) model perturbation, and has complete pre-processing, initial condition perturbation, model perturbation, and post-processing configuration, which can provide a series of probability prediction products; The qualitative and quantitative evaluation of CMA-BJ-EN v1.0 system and its comparison with NCEP global ensemble forecast system show that the system can effectively grasp several major weather processes in North China, and can obtain local refined probability forecast results relative to global ensemble; The statistical results also show that compared with the NCEP GEFS global ensemble forecast, the CMA-BJ-EN v1.0 forecast system can effectively reduce the RMSE of 2m temperature and 10m wind speed, and can effectively improve the probability forecast score of precipitation forecast. The operation of the system can provide users with effective refined probability forecast reference.Keywords: convection permitting, ensemble prediction system, initial condition perturbation, model perturbation